INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Some strains of fluorescent pseudomonads colonize the roots of many crop plants and protect them from diseases caused by soilborne fungal pathogens. Disease suppression by these bacteria involves a blend of mechanisms, including effective competition for nutrients and colonization sites, pathogen inhibition by production of antimicrobial compounds, and induction of resistance in the plant (3, 24, 62). Following introduction into soil, the biocontrol performance of pseudomonads depends largely on their ability to maintain stable populations and to be metabolically active at least over the period needed to exert their beneficial effects. However, in soil these bacteria are exposed to a range of variable biotic and abiotic stress factors, such as competition, predation, and changes in temperature, osmolarity, and availability of water and nutrients (42, 63). Therefore, the sizes of introduced pseudomonad populations may decline considerably within a few weeks, and biocontrol activity often tends to be variable (24, 62).

To ensure survival in changing environments, bacteria rely on regulatory mechanisms that allow them to respond rapidly to stress situations (60). Regulatory elements that make essential contributions to bacterial survival under stress conditions include the alternative sigma factors RpoS (^s) and RpoE (²²; also referred to as AlgU or AlgT in fluorescent pseudomonads). RpoS is required for tolerance of stationary-phase cultures of different Pseudomonas species towards hyperosmolarity, high temperatures, and agents generating reactive oxygen intermediates (ROIs) (23, 52, 61). AlgU contributes to tolerance towards osmotic, oxidative, and heat stresses in the pathogens Pseudomonas aeruginosa (32, 54, 56, 70) and Pseudomonas syringae (26).

The extracytoplasmic function sigma factor AlgU is encoded by the algU gene, which is part of a highly conserved operon in gram-negative bacteria (19, 26, 35, 44). The function of algU has been extensively studied in P. aeruginosa with respect to its roles in stress response and in regulation of biosynthesis of the exopolysaccharide (EPS) alginate. Regulation of AlgU activity in P. aeruginosa is complex. AlgU positively regulates its own transcription (12, 22, 56, 68). Located downstream of algU, the mucABCD genes ensure tight control of AlgU activity in P. aeruginosa (19). The mucA gene encodes a transmembrane protein which acts as an anti-sigma factor for AlgU, and mucB codes for a periplasmic protein which is another negative regulator of AlgU (36, 57, 69). MucC and MucD modulate algU expression, but the precise functions of these proteins have not been fully established yet (5, 6). Binding of AlgU to the inner membrane protein MucA occurs when MucA interacts with the periplasmic MucB protein (36, 44, 50). Environmental stress conditions are thought to destabilize the MucB-MucA-AlgU complex, leading to release of AlgU into the cytosol. As a consequence, AlgU becomes active and transcription of alginate biosynthesis and other genes occurs. Any change in the balance of this regulatory system influences the titer of available active AlgU. For example, mutations in mucA or mucB cause increased activity of AlgU, which leads to mucoidy due to overproduction of alginate (33, 34, 36). Mucoid conversion due to spontaneous lesions in mucA is typically detected in P. aeruginosa isolates from chronically infected cystic fibrosis patients, and the production of copious amounts of alginate is thought to be important for virulence and survival of these bacteria in the lungs (4, 34).

Little is known about the role of AlgU and its negative regulators in plant-beneficial pseudomonads. In the present study, we identified a genomic region which comprises the algU-mucA-mucB gene cluster in Pseudomonas fluorescens CHA0, a well-characterized soil bacterium with broad-spectrum biocontrol activity (24, 65). We found that AlgU, along with MucA, tightly controls EPS biosynthesis and tolerance towards osmotic stress in this bacterium. In contrast to AlgU of pathogenic pseudomonads, AlgU of strain CHA0 does not contribute to survival in response to treatment with heat and ROIs. Finally, we present the first evidence that AlgU is a crucial determinant in adaptation of P. fluorescens to desiccation stress, a major factor that limits bacterial survival in formulations or in soil.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains, plasmids, and culture conditions. The bacterial strains and plasmids used in this study are described in Table 1. P. fluorescens strains were cultivated on nutrient agar (NA) (59), on King's medium B (KMB) agar (27), in nutrient yeast broth (NYB) (59), in KMB broth, and in Luria-Bertani broth (LB) (51) at 30°C with aeration. Escherichia coli and P. aeruginosa strains were grown on NA and in NYB at 37°C. Antimicrobial compounds, when required, were added to the growth media at the following concentrations: ampicillin, 100 µg/ml; chloramphenicol, 25 µg/ml; HgCl₂, 20 µg/ml; gentamicin, 10 µg/ml; kanamycin sulfate, 25 µg/ml; rifampin, 100 µg/ml; and tetracycline hydrochloride, 25 µg/ml for E. coli and 125 µg/ml for P. fluorescens strains. When appropriate, 5-bromo-4-chloro-3-indolyl--D-galactoside (X-Gal) was incorporated into solid media to monitor -galactosidase expression (51).

DNA manipulation and sequencing. Small-scale plasmid DNA preparations were obtained from P. fluorescens strains and pME3049-, pME3087-, and pME3088-based plasmids were isolated from E. coli by the alkaline lysis method (51). All other plasmids were prepared from E. coli by the method of Del Sal et al. (10). Qiagen-tip 100 columns (Qiagen Inc.) were used for large-scale plasmid DNA preparation. Chromosomal DNA of P. fluorescens was isolated as described by Schnider et al. (53). Standard techniques were used for restriction, agarose gel electrophoresis, dephosphorylation, generation of blunt ends with the Klenow fragment of E. coli DNA polymerase I or T4 DNA polymerase (Roche), isolation of DNA fragments from low-melting-point agarose gels, and ligation (51). Restriction fragments were purified from agarose gels with a Geneclean II kit (Bio 101). Bacterial cells were transformed with plasmid DNA by CaCl₂ treatment (51) or electroporation (15). Southern blotting with Hybond N membranes (Amersham), random-primed DNA labeling with digoxigenin-11-dUTP, hybridization, and detection (Roche) were performed by using the protocols of the suppliers. Subclones of pME6202XEV (Table 1; Fig. 1) constructed in pBluescript II KS+ were used for nucleotide sequence determination. Both strands of the 2,825-bp EcoRI-BamHI fragment of pME6202XEV were sequenced by the dideoxy chain termination method using a Sequenase 2.0 kit (United States Biochemical, Cleveland, Ohio) and T7 polymerase from Pharmacia. Nucleotide and deduced amino acid sequences were analyzed with programs of the University of Wisconsin Genetics Computer Group package (version 9.1).

View larger version (21K): [in this window] [in a new window]   FIG. 1.   Physical location of the nadB, algU, mucA, and mucB genes in P. fluorescens CHA0. Symbols: , Tn5 insertion in the chromosome of strain CHA211; , region deleted in strains CHA212 and CHA213M and in plasmid pME6219. The shaded arrows indicate the genes sequenced or partly sequenced on the EcoRI-BamHI fragment of pME6202XEV. The open boxes indicate the genomic inserts in ColE1-based plasmids pME6200B, pME6200X, pME6202X, and pME6202XEV. The lines indicate the fragments cloned into vector pME6000 to obtain pME6217, pME6220, and pME6555 and into vector pME6010B to obtain pME6219, pME6221, and pME6222.

Plasmid mobilization and transposition. Derivatives of the suicide plasmids pME3049, pME3087, and pME3088 were mobilized from E. coli to P. fluorescens with helper plasmid pME497 in triparental matings as described by Schnider et al. (53). Transposon mutagenesis in which E. coli W3110 containing Tn5 suicide plasmid pLG221 (7) was used as the donor strain and P. fluorescens CHA0 was used as the recipient strain was carried out as described previously (37).

Construction of P. fluorescens mutants by gene replacement. For construction of the algU in-frame mutant CHA212 (Fig. 1), the 275-bp NotI-ClaI fragment in algU was replaced by the 56-bp NotI-ClaI polylinker from pBluescript II KS+. The flanking genomic DNA, consisting of the 1.2-kb HincII-NotI fragment and the 1.4-kb ClaI-BamHI fragment of pME6202XEV (Fig. 1), was cloned into suicide vector pME3087 (65). To obtain the mucA in-frame mutant CHA213M (Fig. 1), the 309-bp SmaI-McsI fragment in mucA was deleted. The flanking genomic DNA, consisting of the 0.6-kb PvuII-SmaI fragment and the 0.8-kb McsI-BamHI fragment of pME6202XEV, was cloned into suicide vector pME3088 (65). The derivatives of the suicide plasmids, which carried a tetracycline resistance determinant, were mobilized with helper plasmid pME497 (66) to wild-type strain CHA0. Cells with a chromosomally integrated plasmid were selected for tetracycline resistance. Excision of the vector by a second homologous recombination event was observed after enrichment for tetracycline-sensitive cells (53). The same approach was used to create an algU in-frame mutation in rifampin-resistant strain CHA0-Rif. To obtain the mucA gacA double mutant CHA213M.gacA, plasmid pME3089Km, a derivative of suicide vector pME3088 carrying gacA disrupted by an -Km cassette (29), was mobilized into strain CHA213M. Selection of cells with a chromosomally integrated plasmid and excision of the vector by a second crossover were performed as described above. Selection for kanamycin resistance ensured the presence of the -Km insertion in strain CHA213M.gacA. The algU, mucA, and gacA mutations were checked by Southern blotting (data not shown).

Extraction and quantification of EPS from P. fluorescens. Aliquots (100 µl) of an overnight LB culture of strain CHA0 or one of its derivatives were plated on KMB agar plates. Three replicate plates per strain were prepared and incubated for 24 h at 30°C. Extraction and quantification of EPS were performed by using a procedure described by May and Chakrabarty (39). Briefly, cells were removed from the medium and suspended in 0.9% NaCl. The suspensions were centrifuged for 30 min at 17,700 × g and 4°C to separate the cells from the EPS. The cell pellets were kept for determinations of total cell protein contents. The EPS was repeatedly precipitated and washed with ice-cold absolute ethanol. The contents of uronic acid polymers (components of alginate) in the samples were then assessed by performing the colorimetric carbazole assay (39), using D-mannurolactone (Sigma Chemical Co., St. Louis, Mo.) and alginic acid from seaweed (Macrocystis pyrifera; Sigma) as the standards. The uronic acid content was expressed per milligram of total cell protein as determined by the Lowry method, using bovine serum albumin as the standard.

Purification and biochemical analysis of EPS. Mucoid layers from five KMB agar cultures of highly mucoid mutant strains CHA211 and CHA213M, cultivated under the conditions described above, were scraped off the plates with a sterilized spatula, pooled, and suspended in 50 ml of phosphate-buffered saline (pH 7.2). After vigorous stirring at 4°C for 1 h, the viscous solution was centrifuged at 17,700 × g and 4°C for 4 h to remove the bacterial cells. The remaining proteins in the supernatant were denatured by heating at 80°C for 30 min and removed by centrifugation at 17,700 × g for 30 min. Precipitation of EPS by the addition of ice-cold absolute ethanol, washing, and removal of nucleic acids by digestion with DNase I (type IV; Sigma) and RNase A (type 1A; Sigma) were performed as described by Pedersen et al. (47). The EPS was then further purified by ion-exchange chromatography. Samples were dissolved in 25 mM ammonium carbonate, loaded onto a DEAE-Sepharose CL-6B column (1.6 by 20 cm; bed volume, 40 ml; Pharmacia), which had been equilibrated with the same buffer, and eluted with a linear 25 mM to 1 M ammonium carbonate gradient (39, 47). Fractions (5 ml) containing uronic acids were pooled, extensively dialyzed against sterile water, and lyophilized.

Effect of osmotic stress on algU'-'lacZ expression and growth. P. fluorescens CHA0 and its mutant derivatives carrying an algU'-'lacZ translational fusion on plasmid pME6222 (Table 1; Fig. 1) were grown in 20 ml of KMB without selective antibiotics in 100-ml Erlenmeyer flasks plugged with cotton. Plasmid pME6222 carries the algU promoter region and the 5' end of algU fused to 'lacZ from pMN482 (43). To induce osmotic stress, KMB was supplemented with NaCl (0.6, 0.8, 1.0, or 1.2 M) or sorbitol (1.2 or 1.6 M), a nonionic solute which cannot be metabolized by strain CHA0. For inoculation, aliquots of exponential-growth-phase LB cultures of the bacterial strains were used to adjust the cell concentrations to an optical density at 600 nm (OD₆₀₀) of 0.05. Cultures were incubated with rotational shaking (180 rpm) at 30°C. -Galactosidase specific activities of at least three independent cultures were monitored by the Miller method (51).

Desiccation survival assay performed with filter disks. The sensitivity of P. fluorescens to desiccation on filters was assessed by using a modification of the procedure described by Ophir and Gutnick (46). Dilutions of overnight bacterial cultures were vacuum filtered onto Millipore filters (no. HAWP04700; pore size, 0.45 µm; diameter, 3.5 cm) in order to obtain about 10 to 20 physically separated bacterial cells per filter. The filters were placed onto KMB agar plates and incubated at 30°C for 24 h, which yielded about 5 × 10⁷ CFU per colony that had developed from the individual bacterial cells. Bacterial colonies were then slowly dried by removing the filters from the agar plates, cutting the filters into small pieces so that each piece contained a single bacterial colony, and incubating the pieces in empty petri dishes at 30°C for 24 h. Colonies on filter pieces that were placed on agar medium lacking nutrients (18 g of Serva agar per liter, 1.15 g of K₂HPO₄ per liter, 1.5 g of MgSO₄ · 7H₂O per liter) and were incubated for the same period served as controls. Cells from a single colony on each filter were then suspended in 1 ml of a 0.85% NaCl solution by vigorous mixing with a Vortex mixer for 15 min, and serial dilutions were plated on NA to determine the number of CFU per colony. Washed filter pieces incubated on NA plates showed that almost all cells were removed by this treatment. The level of survival was calculated by determining the percentage of the number of CFU in desiccated colonies relative to the number of CFU in control colonies. Each filter piece was handled separately, and the numbers of CFU were determined for at least five colonies per treatment. The experiment was repeated three times.

Desiccation survival assay performed with soil microcosms. Survival of bacterial strains was monitored in desiccating, sterile, artificial soil and in desiccating, nonsterile, natural soil. The artificial soil consisted of pure vermiculite (expanded with 30% H₂O₂), quartz sand with different sizes of particles, and quartz powder mixed at a ratio of 10:70:20 (by weight) and moistened with 10% (wt/wt) distilled water (25). The artificial soil was autoclaved twice prior to bacterial inoculation. Natural sandy loam soil was collected from the surface horizon of a Swiss cambisol located at Eschikon near Zurich (45). The soil was sieved through a 5-mm mesh screen prior to use, and stones and roots were removed. The bacterial cell suspensions used in microcosms were prepared from exponential-growth-phase cultures grown in LB at 24°C for 16 h. The cells were washed twice in sterile distilled water, and the cell concentration was adjusted to an OD₆₀₀ of 0.2. For inoculation of the artificial or natural soil, 67 ml of the bacterial suspension was thoroughly mixed into 1,000 g of soil with a sterilized spoon to obtain about 10⁷ CFU per g of soil. For natural soil microcosms, rifampin-resistant derivatives of the bacterial strains were used as the inoculants. Aliquots (20 g) of inoculated soil were placed into sterile 200-ml Erlenmeyer flasks with wide openings. The openings were covered with one layer of sterilized paper cloth, which allowed slow evaporation of the soil water. Control microcosms kept at a constant soil moisture level were prepared by placing 200-g portions of sterile, artificial soil or nonsterile, natural soil containing the bacterial inoculum into 500-ml bottles, which then were sealed hermetically. The initial water content of the artificial soil after addition of bacteria was 15.2% ± 0.2%, which corresponded to a soil water potential (_W) of about 0.01 MPa. Natural soil contained 26.9% ± 0.2% water, which corresponded to a _W of about 0.02 MPa. The soil microcosms were incubated in the dark at 24°C with 65% relative humidity. For sampling and enumeration of the bacterial inoculants, the entire contents of a flask with desiccated soil or a 5-g sample from a control soil kept at a constant humidity was suspended in sterile distilled water by vigorous agitation on a shaker at 260 rpm for 20 min. At each time point, the number of cultivable cells and the soil water content were determined for three replicate flasks or soil samples per treatment. The numbers of CFU were determined by plating serial dilutions of the soil suspensions on NA. Rifampin-resistant derivatives were recovered from natural soil by plating samples on NA containing 100 µg of rifampin per ml. No rifampin-resistant background bacterial population was present in the natural soil. CFU data were expressed per gram (dry weight) of soil and were log₁₀ transformed before means and standard deviations were calculated. Water content was assessed by oven drying soil samples at 105°C to constant weight. The _W of soil was determined by using a filter paper method described by McInnes et al. (41).

Susceptibility to oxidative stress. Sensitivity to paraquat (1,1'-dimethyl-4,4'-bipyridinium dichloride; Sigma), hydrogen peroxide (H₂O₂), or sodium hypochlorite (NaOCl) was examined as described by Martin et al. (32). Filter disks (diameter, 6 mm; Millipore) were soaked with 10 µl of paraquat (1.9 or 3.8%, wt/vol), H₂O₂ (3 or 12%, vol/vol), or NaOCl (5 or 10%, vol/vol) and placed on a layer of soft agar (2 ml of NYB with 0.8% agar) containing 100 µl of a P. fluorescens or P. aeruginosa overnight culture covering NA. In another approach, disks (diameter, 10 mm) were soaked with 50-µl portions of the agents mentioned above. The diameters of the inhibition zones surrounding the impregnated disks were measured after overnight incubation at 30°C for P. fluorescens strains and at 37°C for P. aeruginosa strains.

Sensitivity to high temperatures. P. fluorescens CHA0 and mutant derivatives of this strain were grown at 30°C in 20 ml of KMB or NYB in 100-ml Erlenmeyer flasks sealed with cellulose stoppers. When the cultures reached an OD₆₀₀ of 0.5, the flasks were transferred to a water bath and incubated for 0, 5, 10, 20, 30, 60, and 90 min at 42 or 48°C with rotational shaking at 180 rpm. At each time point, three replicate cultures were sampled to determine the number of CFU on NA. CFU were counted after incubation for 24 and 72 h at room temperature. Levels of survival were expressed as percentages of the input number of CFU at time zero.

Nucleotide sequence accession number. The nucleotide sequence of the nadB', algU, mucA, and mucB' genes of P. fluorescens CHA0 has been deposited in the GenBank database under accession no. AF399758.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Isolation and characterization of mucoid mutant CHA211. Following Tn5 mutagenesis of P. fluorescens strain CHA0, 1 of the 1,000 mutants tested had a highly mucoid colony phenotype when it was grown on KMB or NA. In contrast, wild-type strain CHA0 is naturally nonmucoid. The mucoid phenotype of the mutant strain, CHA211, was stable, even when the strain was repeatedly subcultured in nonselective media for more than 1 week. Southern hybridization confirmed that mutant CHA211 contained a single Tn5 insertion (data not shown). The Tn5 insertion was located 550 bp downstream of the initiation codon of the mucA gene (Fig. 1 and see below).

Cloning and sequence analysis of the genes surrounding the Tn5 insertion in strain CHA211. To localize the genomic region mediating mucoidy in strain CHA211, a two-step Tn5-directed cloning strategy (53) was used to clone the wild-type genes corresponding to the genes that were inactivated by the Tn5 insertion in the mutant. Suicide vector pME3049 carrying the kanamycin resistance determinant of Tn5 was mobilized into strain CHA211. Integration of pME3049 into the chromosome occurred at a frequency of 7.2 × 10⁷ per donor. Chromosomal DNA of CHA211::pME3049 was digested with BamHI or XbaI, ligated, and used to transform E. coli DH5. The resulting plasmids, pME6200B and pME6200X, carried 0.8- and 4.5-kb genomic DNA inserts, respectively, downstream of the Tn5 insertion site (Fig. 1). For isolation of the wild-type genes, plasmid pME6200B was transferred into wild-type strain CHA0. Digestion of genomic DNA of CHA0::pME6200B with XhoI, self-ligation, and transformation of E. coli resulted in plasmid pME6202X (Fig. 1) carrying a 14.6-kb genomic fragment. To reduce the size of the insert in pME6202X, the plasmid was digested with EcoRV and ligated. The plasmid obtained, pME6202XEV (Fig. 1), contained a 4.7-kb genomic DNA fragment.

EPS production in P. fluorescens is controlled by the AlgU regulon. Chromosomal algU and mucA in-frame deletion mutations were created in strain CHA0 (Fig. 1) as described in Materials and Methods, and the mutants were tested for EPS production by using the carbazole assay for uronic acids. The nonmucoid strains CHA0 (wild type) and CHA212 (algU mutant) produced very low levels of EPS after incubation on KMB agar for 24 h (Table 2). In contrast, mucA in-frame deletion mutant CHA213M developed a mucoid phenotype on KMB agar and produced copious amounts of EPS (Table 2). The mucA::Tn5 mutant CHA211 (Fig. 1) produced even more mucoid material than the mucA in-frame deletion mutant CH213M produced (Table 2). This may have been due to a polar effect of the Tn5 insertion on other AlgU regulatory genes downstream of mucA, especially mucB (36, 50). The level of EPS production in mucoid mutant CHA213M could be restored to the low wild-type level by complementation with mucA⁺ plasmid pME6219 (Fig. 1), whereas introduction of the cloning vector pME6010B alone had no effect (Table 2). Complementation of the algU mutation in CHA212 by pME6221 carrying intact algU⁺ (Fig. 1) did not significantly affect the nonmucoid phenotype of the strain (Table 2).

Biochemical analysis of EPS from mucoid mutants of P. fluorescens. EPS was extracted and purified from mucoid material of KMB agar cultures of mucA mutants CHA211 and CHA213M by using the procedure described by Pedersen et al. (47). During DEAE-Sepharose ion-exchange chromatography, EPS of both P. fluorescens mutants, like alginate of P. aeruginosa (47), eluted between 0.4 and 0.7 M ammonium carbonate, with a maximum peak at 0.53 M ammonium carbonate. The P. fluorescens EPS preparations had total carbohydrate contents of 62% ± 9% and 87% ± 12% (on a dry weight basis) when seaweed alginate and D-mannuronic acid lactone, respectively, were used as the standards. When either of these standards was used, 46% ± 5% of the dry weight could be attributed to uronic acids, a value which is lower than the uronic acid content (70 to 100%) described for P. aeruginosa alginate (47). Nevertheless, the degree of acetylation of P. fluorescens uronic acids determined by a colorimetric assay was 16% ± 2%, a value identical to the value obtained for P. aeruginosa (47). Acetylation of uronic acid polymers is a typical feature of bacterial alginates, whereas algal alginates are not acetylated (18). No contaminating proteins or nucleic acids were detected in purified EPS from P. fluorescens. These results indicate that mucoidy in mucA mutants of P. fluorescens CHA0 is due to overproduction of an acetylated EPS which may be related to some extent to the EPS produced by P. aeruginosa. Fluorescent pseudomonads have been shown to produce a wide variety of EPS which may differ considerably in composition, and alginate has been detected in only some of these EPS (16, 17).

Kinetics of algU expression in response to osmotic stress. To test algU regulation in response to osmotic stress, we monitored expression of an algU'-'lacZ reporter fusion carried by pME6222 (Table 1; Fig. 1) in wild-type strain CHA0 and mutant derivatives of this strain in the presence of NaCl and the nonionic solute sorbitol. In KMB without NaCl, algU was expressed at low, almost constant levels in strain CHA0 and in algU mutant CHA212 (Fig. 2A). Upon exposure to 0.6 M NaCl, algU expression in wild-type cells was transiently induced about two- to threefold in the early exponential growth phase, whereas no induction occurred in the algU mutant (Fig. 2B). Similar results were obtained when an osmotically equivalent concentration of sorbitol (1.2 M) was added to the growth medium (Table 3). Overexpression of algU from pME6555 or the lack of mucA in strain CHA213M led to expression of the algU'-'lacZ reporter that was significantly greater than the expression in the wild type in the nonsupplemented medium (Table 3). Expression of algU'-'lacZ was further enhanced by exposure to osmotic stress (Table 3). In summary, algU expression in P. fluorescens is induced in response to osmotic stress, and moreover, as in P. aeruginosa (12, 22, 56, 69), AlgU in P. fluorescens positively regulates its own expression.

View larger version (20K): [in this window] [in a new window]   FIG. 2.   Kinetics of expression of an algU'-'lacZ fusion in response to NaCl stress. -Galactosidase activity was assayed in P. fluorescens CHA0 () and algU mutant CHA212 (), both carrying pME6222. Growth of CHA0 () and growth of CHA212 () were determined by measuring OD600. Strains were grown at 30°C in KMB without (A) or with (B) 0.6 M NaCl. The data are means ± standard deviations based on three experiments. Some of the error bars were too small to be included.

AlgU is required for tolerance of P. fluorescens towards osmotic stress. We next tested how different levels of algU expression affect growth and survival of P. fluorescens in response to osmotic stress. To do this, the doubling times of strain CHA0 and mutant derivatives of this strain were determined during exponential growth of the bacteria in KMB supplemented with different concentrations of NaCl or sorbitol. The algU mutant CHA212 was significantly more sensitive to osmotic stress (0.6 or 0.8 M NaCl, 1.2 or 1.6 M sorbitol) than the wild type and the algU mutant complemented with pME6221 (Table 4). After 24 h of incubation, the numbers of CFU obtained from NaCl-stressed cultures of the algU mutant were about 10 to 30 times lower than the numbers of CFU obtained from cultures of the wild type and the complemented mutant exposed to the same stress (data not shown). Constitutive expression of algU from the lac promoter in CHA0/pME6555 made the strain hypersensitive to osmotic stress instead of protecting it (Table 4). Since high levels of AlgU may be toxic to bacterial cells (22, 55), this may have accounted for the increased sensitivity of strain CHA0/pME6555 under stress conditions. Mutant CHA213M lacking the anti-sigma factor MucA exhibited stress tolerance similar to that of the wild type (Table 4). Analogous results were obtained when bacterial strains were exposed to 1.0 M NaCl (data not shown). None of the strains tested was able to grow in medium supplemented with 1.2 M NaCl. In conclusion, AlgU is essential for adaptation of P. fluorescens to osmotic stress, but tolerance towards this stress apparently cannot be improved by overexpression of AlgU.

AlgU is essential for tolerance of P. fluorescens towards desiccation stress in vitro and in soil microcosms. To investigate whether AlgU contributes to tolerance of P. fluorescens towards desiccation stress, strain CHA0 and algU and mucA mutants of this strain were exposed to desiccation stress under various conditions. In the first series of experiments, bacterial colonies grown on filter disks were slowly desiccated for 24 h or were kept at a constant humidity (see Materials and Methods). The numbers of CFU obtained from desiccated colonies of wild-type strain CHA0 were 7% of the numbers of CFU obtained from nondesiccated control colonies. The algU mutant CHA212 did not survive this treatment (<0.1% survival). Complementation of strain CHA212 with algU⁺ plasmid pME6221 restored survival to wild-type levels (9%). mucA mutant CHA213M had a wild-type level of survival (8%), suggesting that a lack of AlgU inhibition and enhanced EPS production do not improve survival in response to desiccation stress.

View larger version (21K): [in this window] [in a new window]   FIG. 3.   Survival of P. fluorescens. Strain CHA0 (wild type) () and its derivatives CHA212 (algU) (), CHA212/pME6010B (algU/) (), and CHA212/pME6221 (algU/algU+) () were included in a desiccating sterile artificial soil consisting of a mixture of quartz sand and vermiculite (A) and in a desiccating natural sandy loam soil (B). For microcosms containing natural soil, rifampin-resistant derivatives of the bacterial strains were used as inoculants. The soil water content () at different times is shown only for the treatment with wild-type strain CHA0 since it did not vary among the different treatments. In the artificial soil, water contents of 15, 10, 5, 1, and 0.5% corresponded to soil W of about 0.01, 0.03, 0.5, 2.7, and 40 MPa, respectively. In the natural soil, water contents of 20, 15, 10, and 1% corresponded to W of about 0.03, 1.5, 3.0, and 9.5 MPa, respectively. The data are means ± standard deviations based on three experiments. Some of the error bars were too small to be included.

AlgU is not required for tolerance of P. fluorescens towards ROIs and heat. AlgU contributes to the tolerance of the pathogens P. aeruginosa and P. syringae towards certain agents that generate ROIs and towards heat (26, 32). To determine whether inactivation of algU or mucA affected the susceptibility of P. fluorescens to ROIs, wild-type strain CHA0, algU mutant CHA212, and mucA mutant CHA213M were exposed to 1.9% (wt/vol) paraquat, 3% (vol/vol) hydrogen peroxide, or 5% (vol/vol) sodium hypochlorite. Table 5 shows that both mutants displayed wild-type susceptibility to these agents. Additional experiments in which the concentrations of the ROI-generating agents were increased up to 10-fold gave analogous results (data not shown). In contrast, an algU mutant (PAO6852) of P. aeruginosa, which was included as a control in the same experiment, was significantly more sensitive to paraquat, but not to H₂O₂ and NaOCl, than wild-type strain PAO1 and mucoid strain PAO568 were (Table 5), confirming previous data of Martin et al. (32). A series of experiments was then performed to evaluate the role of AlgU in tolerance of P. fluorescens towards heat killing. Exponential-growth-phase cultures of strain CHA0 and algU mutant CHA212 were exposed to 42 or 48°C for 5 to 90 min. However, in none of the experiments was there a significant difference in viability between the wild type and the algU mutant (data not shown). In an experiment in which we screened for additional factors that may involve the AlgU-mediated stress response, we found no differences in the tolerance of strain CHA0 and its algU mutant towards a series of other stresses, including exposure to the strong reducing agent dithiothreitol (8 mM), which triggers RpoE function in E. coli (44), exposure to a range of antibiotics, and exposure to pH extremes (data not shown). In summary, these results suggest that AlgU is not required for tolerance towards ROIs and heat in P. fluorescens, in marked contrast to the role of this sigma factor in P. aeruginosa, P. syringae, and E. coli.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In the present study, we identified the algU-mucA-mucB gene cluster in the plant-beneficial strain P. fluorescens CHA0; this gene arrangement is conserved in P. aeruginosa (12, 31, 33) and P. syringae (26). In P. aeruginosa, the algU, mucA, and mucB genes are cotranscribed and encode (respectively) the sigma factor AlgU and its main negative regulators, MucA and MucB, which act in concert to control production of the EPS alginate and the response to extreme environmental stress (19, 32, 50, 54, 57, 70). In P. syringae, AlgU is also a major determinant in the regulation of alginate biosynthesis and stress response (26). Here, we provide evidence that on the one hand, AlgU of P. fluorescens is functionally equivalent to its counterparts in pathogenic pseudomonads with respect to control of EPS production. On the other hand, the role of AlgU in the stress response of P. fluorescens is distinct from its role in the stress responses of other bacteria.

Regulation of EPS production. In P. fluorescens CHA0, EPS production was regulated by at least four mechanisms. First, the level of algU expression was important. This was seen most clearly in a strain with moderate algU overexpression, CHA0/pME6220, which overproduced EPS (Table 2). The level of algU expression is determined by AlgU itself: expression of an algU'-'lacZ reporter was about 40% lower in an algU mutant than in wild-type strain CHA0 (Fig. 2A; Table 3). AlgU autoregulation has also been described for P. aeruginosa (12, 22). The algU promoter region in P. fluorescens CHA0 was found to contain two putative AlgU/RpoE recognition sequences. In P. aeruginosa, the corresponding promoters are designated P1 and P3, and they are absolutely dependent on AlgU (56). Second, the anti-sigma factor MucA made a major contribution to EPS production. Without MucA the EPS levels were 100-fold higher than those in a MucA⁺ background (Table 2). In contrast, the (indirect) effect of MucA on algU transcription was relatively small (Table 3). Third, the GacS-GacA two-component system was essential to sustain high EPS productivity in a mucA mutant (Table 2). This positive effect of the transcriptional regulator GacA was not mediated by AlgU (Table 3). Fourth, the hypermucoid phenotype of strains CHA211 (mucA::Tn5) and CHA0/pME6555 (P_lac-algU⁺mucA⁺) points to negative AlgU control by MucB, as in P. aeruginosa (36, 50).

Role of AlgU in tolerance towards environmental stress. Little is known about the regulatory mechanisms that determine adaptation of plant-beneficial pseudomonads to environmental stress. The present study established that AlgU is a second sigma factor, in addition to RpoS (52), involved in the response of P. fluorescens to extreme environmental stress. Our work shows, for the first time, that AlgU is a key determinant in the tolerance of P. fluorescens towards desiccation stress, since the survival of an algU mutant was severely impaired when the organism was exposed to desiccation in vitro, in a vermiculite-sand mixture mimicking formulation conditions, and in natural soil (Fig. 3). Furthermore, this sigma factor was important in the adaptation of P. fluorescens to high-osmolarity conditions (Table 4), and high concentrations of NaCl or sorbitol stimulated algU expression in both P. fluorescens (Table 3; Fig. 2) and P. syringae (26). Since plant-associated pseudomonads may be exposed to dry and osmotically harsh conditions in the rhizosphere and in the phyllosphere, the capacity to adapt to these conditions is likely to be important for colonization of these plant surfaces (26, 42, 63).